102
131

 

 

 

This is to certify that the

thesis entitled -
Synthesis of the Tetrapyrrolic Macrocycle 3,4,8,9,17,18,22,-

23-Octamethyl-30,31,33,34-Tetraazaheptacyclo[23.3.1.12’5—
.17’10.111’15.116’19.121’243Tetratriaconta-l,3,5,7(33),8,-
10,15(32),16,18,20,22,24(30),25(29—Tridecaene

Presented By

Robert Allan Berger

has been accepted towards fulfillment
of the requirements for

PhoDo degree in Chemistry

gym W
/ Major profess r

 

 

 

Datemmbw 3’ @777

0—7 639

 

SYNTHESIS OF THE TETRAPYRROLIC MACROCYCLE 3,u,8,9,17,18,22,23-
OCTAMETHYL-3O,3l,33,3H—TETRAAZAHEPTACYCLOE23.3.1.12’5.17’10-
.111’15.116’19.121’ZHJTETRATRIACONTA-

1,3,5,7(33),8,10,15(32),16,18,20,22,2H(30),25(29)-TRIDECAENE

By

Robert Allan Berger

A DISSERTATION

Submitted to
Michigan State University
in partial fulfillment of the requirements

for the degree of
DOCTOR OF PHILOSOPHY

Department of Chemistry

1978

ABSTRACT

SYNTHESIS OF THE TETRAPYRROLIC MACROCYCLE
3,“,8,9,l7,18,22,23—OCTAMETHYL-30,31,33,3U—TETRAAZAHEPTACYCLO-

11,15 116,19 121,214

[23.3.1.12’5.17’10.1 ]TETRATRIACONTA-

1,3,5,7(33),8,10,15(32),15,18,20,22,2“(30),25(29)-TRIDECAENE

By

Robert Allan Berger

The synthesis of the next higher member of a family of

vinylogously ring expanded tetrapyrrolic macrocycle 25,26,27,28-

13’6.110’13.115’18joctacosa—l,3,5,—

tetraazapentacycloEZO.2.1.
7,9,ll,l3(27),1A,16,18,20,22(25),23-tridecaene, g, of which
the porphyrins are the first members is described. The con-
densation of 1,3—bis(3,U-dimethylpyrrol—2—yl)cyclohex-l-ene
with 1,3-bis(5—formyl—3,A-dimethylpyrrol-2-yl)cyclohex—l-ene
followed by air oxidation gave 3,“,8,9,17,18,22,23—0ctamethyl—

11,15. 16,19_

30,31,33,3H—tetraazaheptacycloE23.3,1.12’5.17’10.l l

21’2”Jtetratriacenter-J,3,5,7(33),8,10,15(32),16,18,2o,22,2u--

.l
(30),25(29)—tridecaene, {E‘ Both El and its diprotonated salt
were deep green and exhibited an intense Soret-like absorption
at U77 nm. The PMR Spectrum indicated that ii was diatropic

as expected in a [22]annulene. Spectral data are presented

for the structural proof of ET and its precursors.

To my wife, Janet,
who gave constant

faith and encouragement;

To my parents,
Cleva and Calvin Berger,

who never stopped believing

and gave me so much.

ii

ACKNOWLEDGMENTS

I would like to thank the Department of Chemistry at
Michigan State University for providing financial support in
the form of a teaching assistantship for the past four and
one—half years. I would also like to express my appreciation
to Professor Eugene LeGoff for his very patient guidance, his
careful and constructive evaluation of my ideas, and for

arranging financial support through the thin summers.

iii

TABLE OF CONTENTS

Chapter

LIST OF TABLES
LIST OF FIGURES.
INTRODUCTION
PURPOSE OF THIS INVESTIGATION.
INITIAL ATTEMPTS TO SYNTHESIZE PLATYRIN.
SYNTHESIS OF THE PLATYRIN SYSTEM .
EXPERIMENTAL
General Procedure
2,2'-Dipyrrylmethane (Q).

5,5'-Bis(3-oxo-butenyl)-2,2'-
dipyrrylmethane (1e).

5,5'—Bis(3—oxo-propenyl)2,2'-
dipyrrylmethane (ZR).

Attempted Synthesis of 2,2'-
dipyrrylpropane (Q) . . . .

Triphenylmethylfluoroborate (lg).
2,5-Bis(triphenylmethyl)pyrrole (l%).

Dimethyl—N,N'-bis(tosylmethyl)—
glutarimidate (téé) . . .

Dimethyl—N,N'—bis(tosylmethy1)—
thioglutarimidate (IQQ)

N,N'-Bis(tosylmethyl)glutaramide (lg)

N ,N'- Bis(tosylmethyl)thioglutar-
amide (lg). . . . . . . .

N,N'—Bis(glycyl)glutaramide (gl).

iv

Page

vi

vii

l3
15
26
A2
M2
U2

“3

“3

nu
an
AM

“5

“5
U6

H6
“7

Chapter Page

Attempted Synthesis of 1,3-Bis—
(3, A— dicarbethoxypyrrol—2 —yl)~
propane (33). . . . . . . . . . . . . . . . . . M7

3,A—Dimethylpyrrol (33) . . . . . . . . . . . . . H7

3,A,3',u'—Tetramethyldipyrryl-
2,2'-hexacyclotrimethine
tetrafluoroborate (2Q). . . . . . . . . . . . . . H8

1 ,3— —Bis(3, U- dimethylpyrrol— 2- .yl)
cyclohexane (31). . . . . . . . . . . . . . N9

1 ,3- Bis(S- formyl- 3, 4- -dimethylpyrrol-
2- yl)cyclohexane (3%) . . . . . . . . . . . 50
3,u,8,9,17,18,22,23-Octamethyl—30,31,33,3M-

tetraazaheptacyclo[23.3.1.12’5.l7’10.lll’l§

116’19.I21’2u3tetratriaconta-2,U,6,8,lO(33)-
16,18,20,22,2U(30)-decaene (33) . . . . . . . . 51

1,3-Bis(3,u-dimethylpyrrol-2—yl)—
cyclohex-l—ene (3Q) . . . . . . . . . . . . . . . 52

l,3—Bis(5-formyl-3,U-dimethylpyrrol-
2-yl)cyclohex-l—ene (3g). . . . . . . . . . . . . 52

3,“,8,9,17,18,22,23-Octamethyl—30’31,33,3n-

tetraazaheptacycloE23.3.1.12’5.17’lo.lll’15—

116’19.121’2ujtetratriaconta—l,3,5,7,(33)-
8,10,15(32),16,18,2o,22,2u(30),25<29)—
tridecaene (33) . . . . . . . . . . . . . . . . 53
Attempted Complexation of Compound 33 . . . . . . 5A

Reaction Between Compound 33 and boron-
trifluoride . . . . . . . . . . . . . . . . . 55

BIBLIOGRAPHY . . . . . . . . . . . . . . . . . . . . . 56

Table

LIST OF TABLES

Page
Nuclear Magnetic Resonance Absorption
of Bis(trimethylene)octamethylplatyrin
and Related Compounds . . . . . . . . . . . . 3“
Electronic Absorption of Bis(tri-
methylene)octamethylplatyrin and
Related Compounds . . . . . . . . . . . . . . 36

vi

Figure

LIST OF FIGURES

Graph of delocalization energy ye

number of n-electrons as calculated

by HMO, PPP, and SP0 methods.

U'l

IO

Illustration of diamagnetic ring
current effect.

Internal structure of Bis(trimethylene)—
octamethylplatyrin dication

Infrared spectrum of 5,5'-Bis(3—oxo-
butenyl)-2,2'-dipyrrylmethane (1%).
Infrared spectrum of 5,5'-Bis(3—oxo-
propenyl)-2,2'-dipyrrylmethane (ZR)
Infrared spectrum of 2,5-Bis(triphenyl-
methyl)pyrrole (l3)

Infrared spectrum of N,N'-Bis(tosyl—
methyl)glutaramide (3%)

Infrared spectrum of N,N'-Bis(tosyl-
methyl)thioglutaramide (3%)

Infrared spectrum of N,N'-Bis(glycyl)-
glutaramide (2%).

Infrared spectrum of 3,u,3',u'_
Tetramethyldipyrryl-2,2'—hexacyclo—

trimethine tetrafluoroborate (36)

vii

Page

38

59

6O

61

62

63

6A

65

Figure Page

11 Infrared spectrum of l,3-Bis(5-
formyl-3,A-dimethylpyrrol—2-yl)—
cyclohexane (38). . . . . . . . . . . . . . . . 66
12 Infrared spectrum of 3,u,8,9,l7,18,22,23-
Octamethyl-30,3l,33,3A-tetraazaheptacyclo-
[23.3.1.12’5.17’10.111’15.116,19.121’2ul-
tetratriaconta-2,A,6,8,lO(33),l6,l8,20,22,—
2U(30)—decaene (33) . . . . . . . . . . . . . . 67
13 Infrared spectrum of l,3-Bis(5—formyl—
3,U—dimethylpyrrol-2-yl)cyclohex-l-ene (33) . . 68
1a Infrared spectrum of 3,5,3',5'-tetramethyl-
U,u'—diethyldipyrryl-2,2'-hexacyclotrimethine
tetrafluoroborate (33). . . . . . . . . . . . . 69

15 PMR spectrum of 5,5'—Bis(3-oxo-butenyl)—2,2'-

dipyrrylmethane (1g). . . . . . . . . . . . . . 7O
16 PMR spectrum of 5,5'-Bis(3—oxo-propenyl)-

2,2'-dipyrrylmethane (lb).. . . . . . . . . . . 7O
17 PMR spectrum of 2,5-Bis(triphenylmethyl)—

pyrrole (l3). . . . . . . . . . . . . . . . . . 71
I8 PMR spectrum of Dimethyl-N,N'-bis(tosyl-

methyl)glutarimidate (36%). . . . . . . . . . . 71
19 PMR spectrum of N,N'-Bis(tosylmethyl)-

glutaramide (38). . . . . . . . . . . . . . . . 72
2O PMR spectrum of N,N'-Bis(tosylmethyl)-

thioglutaramide (lg). . . . . . . . . . . . . . 72

viii

Figure Page

21 PMR spectrum of N,N'-Bis(glycyl)—
glutaramide (é%)- . . . . . . . . . . . . . . . 73
22 PMR spectrum of 3,U,3',U' Tetramethyl-
dipyrryl—2,2'—hexacyclotrimethine
tetrafluoroborate (26). . . . . . . . . . . . . 73

23 PMR Spectrum of 1,3-Bis(3,U—dimethyl-

pyrrol—2—yl)cyclohexane (23). . . . . . . . . . 7M
2M PMR spectrum of l,3-Bis(5-formyl—

3,u-dimethylpyrrol-2—yl)cyclohexane (38). . . . 7H
25 PMR spectrum of 3,“,8,9,l7,l8,22,23-

Octamethyl-30,3l,33,3u—tetraazahepta-

11,15 16,19 21,2u

cyclo[23.3.1.12’5.l7’10.l l l ]-

tetratriaconta-2,A,6,8,lO(33),l6,l8,20,22,2u-

(30)-decaene (32) . . . . . . . . . . . . . . . 75
26 PMR spectrum of l,3-Bis(3,4—dimethylpyrrol—

2—yl)cyclohex-l—ene (3Q). . . . . . . . . . . . 75
27 PMR spectrum of l,3—Bis(5-formyl—3,A-

dimethylpyrrol-2-yl)cyclohex-l—ene (3g) . . . . 76
28 PMR spectrum of 3,5,3',5'-Tetramethyl-

A,M'-diethyldipyrryl-2,2'—hexacyclo-

trimethine tetrafluoroborate (3i) . . . . . . . 77

ix

INTRODUCTION

Since the early days of organic chemistry the aromatic
compounds have been a source of considerable thought and
constant argument as to their structure and chemical behavior.
Work in the middle of the nineteenth century had shown only
that aromatic compounds generally possessed a high percentage
of carbon to hydrogen, but nothing was determined about their
structures. The ability of aromatic compounds to undergo
substitution reactions while staying inert to most addition
reactions was baffling. In 1865, the first tentative approach
to their structure was given by Kekule in his report that
benzene's structure was satisfied by a six-membered cyclic
molecule of alternating single and double bonds with the re-

1

maining valence used for substitution reactions. However,

his work required further clarification since the number of
isomers in multi-substitution reactions was less than what

2

was predicted by his structure. (See l and 2). He proposed

that the double and single bonds interchanged rapidly such

that it is possible to separate only the reported number of
isomers. Many other structures proposed for benzene dis-
appeared quickly since they failed to explain experimental
data. That double and single bonds might be able to inter-
change was not considered a problem since l,H-addition to
conjugated systems was known. Although why they behaved

so was a mystery.

3

Further clarification was presented by Thiele when he
proposed the concept of partial bonds existing when double
bonds were in conjugation. He proposed that in benzene
there were no discrete single or double bonds in the ring
but six partial bonds that were equivalent. Conjugation
would be a mixing of bond character. Benzene and aromatic
compounds are the extreme case giving total equivalency,
while l,u—addition across butadiene gives partial mixing

of bond character. This approach was not accepted since it
predicted that all cyclic compounds containing alternating
double bonds should be aromatic. The subsequent synthesis

A showed that there must be

of cyclooctatetraene in 1913
something else in the definition of an aromatic compound.

Since aromatic compounds containing a large number of carbon
atoms were unknown, a gentle nudge was required from outside
work before anything more could be accomplished. This nudge
came with the discovery of the electron. Armit and Robinson5

proposed that the electron was the particle most likely

responsible for the concept of valency, and that electrons

in aromatic molecules somehow behaved similarly to those
in closed shells of the noble gases. This brought back the
older idea of bonding equivalency and was used to give a

new structure for benzene; i.e., a hexagon enclosing a circle,

3.

All the work to this time was useful, but nothing was
proposed as to why the "aromatic sextet" was so unique.
The mathematics of quantum mechanics became the next step.

6

The valence bond theory by Heitler—London and the molecular
orbital theory by Hfickel7 tried to explain aromaticity on
the basis of ground state energies versus transition state
energies that occurred during reactions. These two theories
were the first to explain why aromatic reactions were so
different from alkene reactions. Hfickel also predicted what
new compounds could be expected to exhibit aromatic character.
Both theories worked with varying degrees of success, yet this
was considered reasonable since both were designed with dif-
ferent assumptions.

In valence bond theory, a molecule is regarded as a

group of atoms being brought together while still retaining

a large degree of their individual character. A specific

valence bond is considered to involve only two atoms, and each
bond is based on the overlap of atomic orbitals. Then the
symmetry of aromatic molecules requires explanation by the
concept of "resonance". Whenever more than one molecular
structure can be drawn for a molecule and the only difference
is the location of the electrons, that molecule is said
to possess resonance. Structures 3 and g are resonance pic—
tures of benzene since six electrons in three bonds are
shifted to give different structures of benzene. The dif-
ference between the theoretical energy of a particular reson-
ance structure and the experimentally determined energy is
called the resonance energy. For benzene, the resonance
energy is determined by comparing its energy to that of three
isolated olefinic bonds.

A major shortcoming of the valence bond theory is that
it predicts a stabilizing resonance energy for cyclobutadiene

8

and cyclooctatetraene. Since this has not been observed,
its use as a predicting theory has been tarnished. While
not completely correct in its interpretation of aromaticity,
the valence bond theory has been of major importance. Its
approach uses structures that the chemist is more familiar
with and hence, easily understandable.9
The molecular orbital (MO) theory, as developed by

Huckel, treats bonding as bringing the orbitals of two or
more atoms together to form a linear combination of atomic

orbitals which produces an equal number of molecular orbitals

of stable and unstable energies. The electrons are placed

in molecular orbitals filling the lowest energy orbitals
first and each succeeding orbital, according to Hund's Rule
and the Pauli Exclusion Principle, until all the electrons
have been used. While not as easy to visualize as the com-
monly accepted two-electron bond, the predictive power of
HUckel's MO theory led to his discovery of a general rule
which is known as Hfickel's Rule. It states that "amongst
fully conjugated, planar monocyclic polyolefins only those
possessing (Mn + 2) n-electrons, where 8 is an integer, will

6 As a corollary, all similar

have special aromatic stability."
systems containing An w—electrons will not possess any special
aromatic stability and should be unstable if all bonds are
equivalent. This rule has been found to work for many com-
pounds.

Large, fully-conjugated, monocyclic ring systems have
shown that Hfickel's Rule and MO theory are not valid beyond
a certain size. As n gets larger MO theory predicts that
the resonance energy should constantly increase, while the
energy difference between (“n + 2) and An n-electron systems
should approach zero (see Figure 1).10 Only the latter has
been shown to be true for the larger annulenes. Beyond a
22 n-electron macrocycle, some predicted aromatic annulenes

11 while [22] annulene

have been shown to act as polyolefins,
is still aromatic as indicated by PMR studies (see below).12
Further, the delocalization energy of the annulenes does

not increase with size. MO theory also predicts resonance

stabilization for even the An series, which has not been

 

 

 

 

 

415

L L 1 l J l l_ L 1 JE

4 6 8 10121416182022242628

 

Figure 1. Graph of delocalization energy ye number of n-
electrons as calculated by HMO, PPP, and SP0
methods.

reported. The smaller members of the An series have been
shown to be destabilized, when planar, which means a negative
stabilization energy.l3 On this basis, MO theory gives good
qualitative predictions for determining aromatic compounds,
but is still a poor method for obtaining quantitative values
for delocalization energies.

Since 1931 many other theories have been proposed which
tried to remove inconsistencies such that better qualitative
values could be obtained using fewer or more logical approxi-

mations in the calculations. In 1965, Dewar10

published
some results where he used three methods to determine the
delocalization energies of the annulenes; i.e., the simplest
of the conjugated ring systems. The three methods, as seen

in Figure l, are the Hfickel MO theory; the PPP method, a

combination of Pople's method with Pariser and Parr values;
and the SPO or split p-orbital method. The latter two methods
compensate for some of the approximations used in the MO
theory; i.e., the assumptions of a planar polygon geometry,
equal carbon—carbon bond lengths, and no electron-electron
repulsions. The MO results are useful, but the two new
methods have shown more reasonable agreement with available
experimental data. All three methods agree that relative
stability energies approach a constant value as 2 gets larger.
However, the PPP and SP0 methods are consistent with experi—
mental data showing there is no stabilizing energy when n
becomes too large a number. The PPP and SPO methods do
account for the negative delocalization energies of anti—
aromaticitylu of the Mn series and show at approximately what
value of n the (Mn + 2) system becomes a non—aromatic poly-
olefin. Experimental observations of [22] annulenelz show
it is aromatic while a [26] annulenell is non—aromatic.

Based on his calculations Dewar proposed a definition

for aromaticity that stated:15

Cyclic conjugated systems

are considered aromatic if cyclic delocalization of electrons
makesa.negative contribution to their heats of formation.
Dewar's work has been confirmed by Figeysl6 even though

there is still disagreement with experimental work on how
rapidly the resonance energy decreases for the (Mn + 2)
systems as Q gets larger.

A major problem that existed with the Dewar and Hfickel

works was how to determine quickly if a compound was aromatic.

 

Their works used resonance energy for the definition of
aromaticity. Two tedious experimental methods, heats of
combustion and hydrogenation, were the only ways to verify
aromaticity. Fortunately, a much simpler method exists

17-19

through the diamagnetic anisotropy of aromatic molecules.
The major benefit of this property is that paramagnetic and
diamagnetic effects may be observed through chemical shifts
in the nuclear magnetic resonance of protons (PMR) on the

20 Such an effect is ob-

exterior or interior of the ring.
served when an aromatic molecule in the ground state is
perturbed very slightly in the presence of a magnetic field.
The principle requires a closed conjugated path where
the n~electrons would be free to circulate around the mole-
cule. If an external magnetic field, HO, Vfifis then added,
a "ring current" would be induced due to the flow of n-elec—
trons on the circular path. The "ring current" would in turn
produce its own small magnetic field, H'. If the induced
current is perpendicular to Ho’ then H' would oppose HO and
create shifts that would be discernable by PMR. (See Figure
2).
As illustrated in Figure 2, a proton external to the ring
feels a positive enhancement of the applied magnetic field
by the induced magnetic field. The net effect on this proton
is that it is shifted to a lower field than what is found

for an isolated proton. Conversely, a proton inside the

ring is shifted to a higher field since the induced magnetic

Induced
current

 

 

~

Induced}
flew

Figure 2. Illustration of diamagnetic ring current effect.

field is opposed to the applied field. Such a phenomenon
is due to a diamagnetic ring current and molecules which

exhibit this property are called diatropic.21

The opposite
effect, where external protons are shifted upfield and inter—
ior protons shifted downfield, is due to a paramagnetic ring
current and molecules which exhibit this effect are called
paratropic.21 A third term, atropic, is used to describe
molecules in which no ring current effects are seen. Earlier
use of the term "non~aromatic" has been replaced by para-
tropic and atropic since the two effects are different and

it is necessary to differentiate between the two forms.
Currently, a molecule which possesses diamagnetic anisotropy
cannot give a quantitative value to the degree of aromaticity

through PMR, but as long as protons are available on the

10

ring this effect is an excellent qualitative determination
of n—electron delocalization.

Sondheimer has made extensive use of this phenomenon
for the determination of n-electron delocalization in the
annulenes. Since these molecules contain a large number of
internal and external protons in differing numbers, they
are well suited to this technique. Sondheimer has determined
the PMR spectra for a series of [An] annulenes and [Mn + 2]
annulenes and found them to be paratropic and diatropic,

11’22 Some of the larger annulenes were also

respectively.
studied and these gave chemical shifts of isolated olefinic
protons of atropic character. Since this technique can also
be used to detect diatropism in hetero—aromatics, e.g., por-
phyrins, it is a simple method for determining if a molecule
can possess a ring current and, therefore, be aromatic.
Another method for the determination of aromatic character
is through study of the uv-visible spectra of aromatic mole-
cules. In Huckel's MO theory, aromatic compounds have
degenerate highest filled and lowest unfilled orbitals. The
absorption of light in the visible and ultraviolet region cor-
responds to electronic transitions from lower to higher energy
states. When the number of possible transitions between the
highest filled and lowest unfilled orbitals is determined for
an aromatic compound, its spectrum may be predicted as to
wavelength and relative intensity of each band. An interest-

ing development is that for a series of related aromatic

11

compounds the uv—visible spectra can possess the same general

23 If the spectrum has been obtained for one aromatic

shape.
compound, e.g., benzene, then a larger aromatic homolog, e.g.,
[l8] annulene, may have a spectrum that only differs in its
shift to a lower energy. Consequently, the spectrum of a new
compound, when compared to a known aromatic compound's spec—
trum, can be used to determine whether the new compound is a
homologous molecule, i.e., aromatic. The use of UV-Visible
spectrophotometry to discover aromaticity is not as quali-
tative as PMR spectrometry, since substitution effects may

be large. Together, the two methods can be useful techniques
for the determination of aromaticity in new molecules.

From this brief history of aromatic compounds, the
arguments for determining just what is meant by aromaticity
have been presented:

(1) A high percentage of carbon to hydrogen led to alter-

nating double bonds;

(2) Equivalent bonds were explained by rapid inter—
changing, then partial bonds, then a magical
"aromatic sextet";

(3) Quantum mechanics gave valence bond theory, mole-
cular orbital theory, Huckel's Rule, the PPP method,
and SP0 method.

The expansion from benzene to benzenoid to non—benzenoid

species has broadened the boundaries of the definition of

aromaticity to what is now used. Also, the determination

 

12

of aromaticity has been simplified. Early use of heats of
combustion or hydrogenation has changed to the use of PMR
chemical shifts for proving the predicted diatropism or

paratropism from calculations proposed earlier.

PURPOSE OF THIS INVESTIGATION

Sondheimer has determined that n—electron delocalization
is detectable by PMR spectrometry for a large number of
annulenes and dehydroannulenes. He also found a 22 n—electron
system to be diatropic,l2 while a 24 w—electron system was

paratropic.2u There is some argument as to whether or not

11’25’26 A major problem

a 26 n-electron system is aromatic.
with many annulenes is that aromaticity can be temperature
dependent, due to flexibility in their rings. It is possible
that no diatropic effect was seen, since the molecules were
in a conformation that was unfavorable to a ring current.
This investigation was proposed to produce a system that
would be rigid enough to show if the Sondheimer results were
viable for all macrocyclic molecules.

Perhaps the best known aromatic macrocyclic system is
that of the porphyrins, Q. Porphyrins are very stable l8

n—electron molecules, incorporating four pyrrole rings in

8.1:

13

1U

their structures. If the meso-bridges could be expanded from
one carbon to three or five carbons, this system could prove
ideal for the rigid structure required in the proposal.
Several methods have been devised for extending the meso-
bridges,27"29 and this investigation tried to use them

in the preparation of a new macrocyclic compound. Since

new territory was being explored, it was decided to make

the next larger homo-porphyrin that should prove to be di-
atropic; i.e., a 22 w-electron system. This molecule would
possess alternating trimethine and methene meso-bridges.

We propose to call this the platyrin* system, 5.

 

*From the Greek word "platys", (meaning broad or wide) and

the -rin suffix (from porphyrin).

INITIAL ATTEMPTS TO SYNTHESIZE PLATYRIN

If classical porphyrin syntheses were to be followed,
two main ways of building the platyrin system would be sug—
gested:

(1) Build the 2,2'-dipyrrylmethane, 6, and place three—

carbon handles on the free a—positions for use in
a later condensation with another molecule of 6

to form 3 (3%, X = CH3; 3Q, X = H); (see Scheme 1)

Scheme 1’
X
\\ \\ \ \0
NH NH
_4>
NH NH Zt’R
/ / o
/ / / /

15

l6

(2) Build the 2,2'—dipyrrylpropane, 8, place one-carbon
handles on its free a—positions and condense with
another molecule of 8 to form 3. (See Scheme 2)

Scheme 2

g \\ /
NH HN

 

 

Each of these schemes required the oxidation of the newly-
formed macrocycle to the platyrin. We decided to use air in
each case, since many porphyrin syntheses were successful
when air was used as the oxidizing agent.

We decided to attempt Scheme 1 first, since there existed

an easy preparation of 6 from pyrrole, 9.30’31 (See Scheme 3)

473—»

t
o
\ / \ /

Scheme 3

 

l7

Pyrrole was nucleophilic enough to react with thiophosgene
to give 2,2'—dipyrrylthione, lg. Thiones, when treated with
basic peroxide, gave ketones. This method was used to make
2,2'—dipyrrylketone, 33, The reduction of kl was achieved
through sodium borohydride and morpholine treatment to give
6.

The three—carbon functional groups were also easily
synthesized since pyrroles with free a-positions undergo a
Michael reaction with a,B—unsaturated carbonyl compounds.32
Two methanes were synthesized (1a, X = CH3; 32, X = H) using
3-butyn-2-one and propargyl aldehyde.33

It was at this point that Scheme 1 failed, since the
final condensation to form the macrocycles, 3a and 38, required
an acid catalyst. It had been hoped that the condensation
would occur more rapidly than any side reactions. Unfor-
tunately, a mineral acid was required, and in all cases, the
decomposition of the existing methane bridge was the first
reaction. This was followed by rapid polymerization of the
unprotected pyrroles. The characteristic chromophore of a
dipyrryltrimethine at 580 nm could be observed, but no
material could be isolated which could be identified.

A curious side reaction of this scheme occurred while
searching for a strong acid that might be used for the
condensation. We proposed to try a compound which is not

generally used as an acid, triphenylmethyl tetrafluoro-

borate, 13. A model system was set up under anhydrous

 

conditions where 2—pyrrylaldehyde and pyrrole were treated
with 13 in acetonitrile. A colorless crystalline preci-
pitate was immediately formed as the only recoverable com—
pound. It was identified through CMR spectrometry as 2,5—
bis(triphenylmethyl)pyrrole, l3. Triphenylmethyl—2—pyrrole,
13, has been made in quantitative yields from triphenyl-

carbinol and pyrrole in refluxing acetic acid,3u but this

.W<~ ~

©
©cflc©
O;

was the first reported synthesis of the bis-adduct. None
of compound 33 was isolated from the reaction. A second
reaction was set up using only pyrrole, and 33 was the
product again. No reaction occurred with 2—pyrrylaldehyde.
Although the methane bridges in la and ZR seemed to be
too sensitive to work with, it was still desired to use
pyrrole as our starting material. We hoped that a dipyr—
rylpropane would be more stable to acid. Consequently, we
switched to Scheme 2 and set out to make 8. Three-carbon

bridges had been made by condensing pyrroles and

l9

1,3,3—triethoxypropene,29 to give dipyrryltrimethines. If
the dipyrryltrimethine could be made from pyrrole and re-
duced, compound 8 would be made. Each time the reaction
was tried, it would turn deep blue, as is characteristic of
dipyrryltrimethines, but the only product obtained was an
insoluble tar. For the reaction to proceed, a heavily sub-
stituted pyrrole was necessary or the product quickly de-
composed.

Once again a new plan of attack was needed. All the
pyrroles used had been unprotected on their 8—positions, and
we felt that this might have contributed to the poor yields
of the desired products. While it is known that the a—posi—
tions of the pyrroles are more nucleophilic than the B—posi—
tions, the difference can be very slight and hard to control.
Therefore, a good supply of B—protected pyrroles was required
to eliminate reactions at the B-positions. If these pyrroles
were not as acid—sensitive as unprotected pyrroles, a modi-
fied scheme, similar to Scheme 2, could be used to make a
substituted platyrin. Van Leusen gave a source of 3,U—di—
substituted pyrroles, when he reported a new general synthesis
using tosylmethylisocyanide (TosMIC).35 Under basic condi-
tions, TosMIC reacted with a,e-unsaturated ketones, esters
or nitriles to give 3—acyl-pyrroles, pyrrole-3—carboxylates
and 3—cyanopyrroles, respectively. (See Scheme A) The only
problem with this approach was that these pyrroles were un—

symmetrical. It would be very difficult to synthesize just

 

2O

 

Scheme A
TOS’CH2-N=C
.+
R'-CH=CH-9-R2
0

(Where R1 = aryl, alkyl, or H; R2 = aryl, alkyl, or alkoxide)

one product in further reactions, since there were two non-
equivalent, reactive a-positions on each pyrrole.

36

A later paper by van Leusen apparently solved this
nonequivalency problem when he reported the synthesis of
2,3,U—trisubstituted pyrroles using N-tosylmethylimino com-

pounds, 33.

- .. : "R
1'08 CHZN C‘X

it

(Where R = alkyl, aryl; X = OCH3, SCH3)

The major advantage to this method was that the functional
group on the 2-carbon could be an alkyl group such as pro-
pane. We proposed to make a bis-adduct connected by a three-
carbon chain. After further reaction, the product would be a

1,3-dipyrrylpropane. We wanted N,N'-bis(tosylmethyl)—

21

glutarimine, 36%,b, to react with an a,B—unsaturated ketone.

This would give a 1,3-bis(3-acyl-H—alkyl-2-pyrryl)propane,

, x x. _

NM

(For téé: X = OCH3; for ték’ X = SCH3)

33. Using this approach, only one dipyrrylpropane would be

synthesized.

 

t1

Sodium p-toluenesulfinate, formaldehyde and glutaramide
were condensed together in acid to give N,N'-bis(tosylmethyl)-
glutaramide, l8. Compound l8 was then O-methylated with
methyl fluorosulfonate (Magic Methyl) to produce dimethyl—
N,N'-bis(tosylmethyl)glutarimidate, 16g. (See Scheme 5)

Compound 16% was believed to be a mixture of syn and anti

22

Scheme 5

9 9
HZNCCHZCHZCHZ CNHZ
0 i o

n u %8
TosCHzNHC CHZCHZCHZCNHCHZTOS

i
,OCH3 CH3O\

TosCH N=C
2 \ /
CHZCHZCHZ

C=NCH2Tos ma

isomers, since the PMR spectrum showed two unequal singlets
at 63.93 and 3.88 for the methoxy hydrogens. Attempts to
36

make 33 from lée were unsuccessful. Van Leusen reported
success using only methyl groups for "R" in Compound 33 in
making a pyrrole. However, he was able to place a phenyl
ring on the 2-carbon of a pyrrole if the reaction was run
through a thioimine intermediate. Following this approach,
Compound 33 was converted into N,N'-bis(tosylmethyl)thio-

glutaramide, 32, using P285 in p-dioxane. However,

23

S—methylation with Magic Methyl gave no identifiable material.
Apparently, steric hindrance kept the pyrroles from forming,
or the requirement of two reactions on the bis-adduct kept
the yields too low for further work.

A second method for making l,3—dipyrrylpropanes with the
B-positions blocked was attempted through the use of A2—
oxazolium—S—ones, €8,37 which were made by dehydrating amino

acids. Such a synthesis would have proceeded as in Scheme 6.

(Where R = alkyl)
N-(R

2O
’Vb

Glycine and glutaroyl chloride were condensed to form N,N'-
bis(glycyl)glutaramide, 33. If 33 was heated gently in
acetic anhydride it should have formed Compound 33. However,
compounds like 2Q are highly reactive, and when synthesized,
they are usually converted into another product, in_§i§u.
With dimethylacetylenedicarboxylate, a l,3—dipolar addition
to 33 was possible, and this should have given l,3-bis(3,u—
carbethoxypyrrol-2-yl)propane, 33. None of Compound 33 was
ever isolated.

Even with a deactivated pyrrole system as found here, a
Michael-type addition by the nitrogen on a,8-unsaturated

carboxylates has been reported38 to occur when the pyrrolic

 

2“

Scheme 6

o
HOZCCHZNHCCHZCHZCHZCNHCHZCOZH

a
, o o
0-1M ‘70

N N
.212,

cozcu3 COZCH3

"36°25 \ / COZCH3

NH rm

3%

nitrogen is unprotected. Only a phenyl group on the a—posi-
tion has prevented further reaction by the pyrrole.39 The
PMR spectra of the l,3—cycloaddition reactions had always
shown at least twice as many methyl groups as were desired.
This led us to believe that a Michael—type reaction had
occurred.

Since the new methods were unsuccessful for synthesizing

a l,3—dipyrrylpropane directly, another approach was tried.

25

First a 3,4-disubstituted pyrrole would be made, and then,

the three-carbon linkage would be added between two pyrroles.
Unless a symmetrical 3,H-disubstituted pyrrole could be easily
made, the low yields inherent to unsymmetrical pyrrole re—
actions would have to be expected for the overall synthesis

of platyrin.

SYNTHESIS OF THE PLATYRIN SYSTEM

Ideally, the pyrrole for the synthesis of the platyrin
system should have both B—positions blocked with symmetrical
functional groups that do not significantly reduce the re-
activity of the unsubstituted d-positions. A compound similar
to a 3,A-dialkylpyrrole was desired for the synthesis.
Early in 1976, Ichimura, a $1.1”) reported the synthesis of
3,u-dimethylpyrrole, 33, in good yields (3u_usz) through a

thiazine-l-oxide intermediate, 33. When 33 was treated with

CH3

CH3 8

a strong base, 33 was isolated. Our group duplicated this
work, since 33 seemed appropriate in our synthesis. However,
some changes were necessary. When strong base was added
to 33, the reaction was explosive and our yields were less
than reported. A large excess of base was reported as neces—
sary for good yields. We believed that 33 added to the base
would fulfill this requirement. When this modification was
tried, the reaction was much calmer and yields were increased
to 50-55%.

Even with B—positions blocked, a trimethine formed from

33 and 1,3,3-trimethoxypropene might not be rigid enough

26

for later steps in the overall synthesis of platyrin. Con—
sequently, we decided to make a bulky trimethine from the
condensation of 33 and l,3—cyclohexanedione.27 This trimethine
was 3,u,3',A'-tetramethyldipyrryl~2,2'-hexacyclotrimethine

tetrafluoroborate, 36. Compound 36 was a blue, crystalline

\ \
H

\
H N BF
N O 4

32

solid that could be characterized by PMR spectrometry. Some
confusion occurred on the structure identification of 36,
since the PMR location of the proton on the trimethine bridge
was very solvent-dependent. The proton signal shifted
through the a—pyrrolic hydrogen signal when the solvent was
changed from deuterochloroform (CD013) to deuterated dimethyl-
sulfoxide (DMSO—d6). In CDCl3, the peak was at 58.u ppm

while in DMSO-d6, the peak was at 67.2 ppm.

Since 36 now existed, it could be reduced and Scheme 2
would be viable again. A reduction of 36 was necessary since
the conjugated system of a trimethine makes the d—positions
of the pyrroles non—reactive to electrophilic attack by the
reagents needed for further synthesis. However, a reduction

with catalyst and hydrogen would remove the conjugation and

reactivate the deposition on the pyrrole rings. The

28

reduction of gé gave l,3-bis(3,u-dimethylpyrrol—2-yl)cyclo—
hexane, g1. Compound 31 was characterized by PMR spectrometry
as a l,3—disubstituted cyclohexane. Compound 51 was air-
sensitive, as shown by a rapid color change when exposed to

air. Therefore, 31 was always converted into another molecule

NH HN
£1

which was more stable. Since it was now possible to place
functional groups on the remaining a-positions, g1 was usually
converted into the next compound of Scheme 2. In this case,

a formylation reaction with POCl3/DMF was used to make, in
81% yield: l,3-bis(5-formyl-3,u-dimethylpyrrol—2—yl)cyclo-

hexane, fig. The PMR spectrum of 28 showed a single peak at

\ /

\ NH HN
CHO CHO

%§

69.33 ppm for the two equivalent aldehyde hydrogens.

29

Many porphyrins have been synthesized by condensing a
5,5'-diformyl—2,2'—dipyrrylmethane with a 2,2'—dipyrrylmethane
in acid solvents. The porphyrin formations were then com-
pleted by air oxidation. Generally, the air was not used
at the beginning of the reaction, since dipyrrylmethanes may
be oxidized to dipyrrylmethenes. When oxidized, they are
not as active to condensation reactions. Therefore, an oxi-
dizing agent was not wanted or required until all the condensa-
tion reactions were completed. We proposed to use a similar
reaction with compounds ax and 3%. First, the condensation
would occur in acid solvent and then air would be bubbled
into the solvent to perform the oxidation. When this re—
action was run, the product showed the air oxidation did not
occur, although the condensation did and a sample was collect—
ed of the macrocycle: 3,4,8,9,17,18,22,23-octamethyl—30,3l—
33,3M—tetraazaheptacycloE23.3.1.12’5.17’10.lll’15.116’19.121’2u]—

tetratriaconta-2,M,6,8,lO(33),16,18,20,22,2U(30)—decaene, £2.

£2

 

 

30

Compound g2 had the same structure as a platyrin except
it lacked six oxidized carbons. The structure of compound
g2 was determined by PMR spectrometry to be symmetrical and
did not show any signs of a ring current. The mass spectrum
of compound g2 gave a maximum m/e value of 560, which cor—
responded to the parent peak. If the oxidation had occurred,
the parent value of the platyrin should have been 55”. Also,
the uv-visible spectrum showed a single broad band at 429
nm, which is characteristic of neutral dipyrrylmethenes.

The electronic spectrum of a platyrin was expected to be
much more complex in the visible region. There was some evi—
dence in uv—visible spectra of some material with a more
complex spectrum existing in the precipitated samples of 32.
Since only trace amounts existed, none of this particular
compound could be isolated as a solid. Solutions could be
obtained which possessed this complex spectrum, and it was
later proven to be the platyrin by comparison of uv-visible
spectra with a genuine sample. Since no sample of this com—
plex spectrum could be isolated as a solid for further
characterization tests, we decided to try a variation of
ifldesynthesis.that led to £2. We felt it was too difficult
to oxidize a propane with oxygen.

We wanted two 3-carbon bridges to be oxidized. In a
porphyrin synthesis, only one-carbon bridges were oxidized
by air. To make our synthesis more like the construction

of a porphyrin, we decided to reduce partially the trimethine

 

31

linkage in compound gé. This way, only one carbon at each
of the four meso-bridges would need to be oxidized to complete
the formation of the platyrin. Also, the intermediate, like
g1, should still be reactive at the a-positions of the pyr-
roles, since most of the conjugation in the trimethine would
have been reduced.

The partial reduction of gé used a mild reducing agent.)41
A solution of gé could be quickly reduced with sodium boro-
hydride and leave as the only product: l,3-bis(3,u-dimethyl—
pyrrol-2-yl)cyclohex-l-ene, @Q. When a PMR spectrum was taken
of éQ: its vinyl peak was at 65.7 ppm. This gave a possible

explanation for how samples of £2 contained a small amount

of the platyrin. When g1 was made earlier, some of its PMR

«39,
spectra contained a small extra peak at 56.1 ppm. This
generally happened for short reaction times or low hydrogen
pressure in the reduction. We believe that @Q was synthesized
in the reduction as an intermediate and not all of it was
converted to gz. When the cyclization reaction was run an
oxidation intermediate, between 32 and the platyrin, might
have been formed which could be much easier to air oxidize

to the platyrin. Such a compound would have a structure

32

like @l. No compound like él was characterized, but that was

 

expected since 3% should have been converted to the platyrin.
Indeed, if a molecule like gl could be air-oxidized, our new
intermediates might work.
Trying to keep overall yields as high as possible, we
decided not to react go and 2Q to form él. Instead, £0 was
’b ’b ’b ’b
formylated with POCl3/DMF to synthesize an 80% yield of

l,3—bis(5—formyl-3,u-dimethylpyrrol-2-yl)cyclohex-l—ene, §§°

CH0 CHO
@%

33

The PMR spectrum of 3g showed two equal singlets at 69.N0 and
9.35 ppm for the aldehydes. This was reasonable, since with
the double bond in 3%, it was expected that the two aldehydes
should not be equivalent. The methyls on the two pyrroles
were also not equivalent in the PMR spectrum. Now we felt
that the final condensation could be attempted to synthesize
a platyrin.

Condensation of 39 and 3% in HBr/methanol, with air
bubbled into the solution, gave a small amount of a greenish
material with a melting point greater than 300°C. We assigned
the following name to this compound: 3,H,8,9,l7,l8,22,23-
octamethyl—3O,31,33,3U-tetraazaheptacyclo[23.3.1.12’5.l7’10-

.111,15.116,19_121,2u

Jtetratriconta-l,3,5,7(33),8,10,15(32)—
l6,l8,20,22,2U(30),25(29)-tridecaene, 33. Its trivial name
is bis(trimethylene)octamethylplatyrin. At room tempera—
ture, the PMR spectrum showed a very strong ring current
effect. This was consistent with 3% being an aromatic
macrocycle. (See Table l). The two one-carbon meso-bridges
were external to the ring and they were shifted down-field
to 611.6u ppm. The two protons located on the middle carbon
of the three—carbon meso-bridges were in the interior of

the ring and they were shifted up—field to 6-8.97 ppm. The
PMR spectrum was taken in CDCl3-CF3CO2H so that all four
nitrogens were protonated. These acidic protons were

equivalent and shifted up-field to 6-5.6 ppm. The values

obtained for the exterior meso-protons and the interior

3U

 

Table 1. Nuclear Magnetic Resonance Absorption of Bis(tri-
methylene)octamethylplatyrin and Related Compounds

h2,a

 

 

 

 

Compound CH3 Meso CH NH
Bis(trimethylene)octa-b u.22 11.6ug -5.6
methylplatyrin bis—tri- “.17 ~8.97
fluoroacetate salt
Octamethylporphyrine 3.76 10.98 -u.82
bis—trifluoroacetate
salt
Decamethylsapphyrine “.31 -6.15
bis-trifluoroacetate “.15 -6.A6
salt -6.73
aAs d-values referred to internal tetramethylsilane. bTaken

in deuterochloroform with trifluoroacetic acid.

ring. dInterior of ring.

cExterior of
8Taken in trifluoroacetic acid.

35

nitrogen protons of the platyrin are consistent with values
that have been determined on similar compounds such as octa-
methylporphyrin and decamethylsapphyrin.“2 The Spectrum of
the neutral platyrin was not obtained since it was not solu-
ble in deuterated solvents in which it was stable. Treatment
of 33 with acidic deuterated solvents gave rapid proton ex-
change of all meso and nitrogen protons.

The uv—visible spectra were taken for both the "free
base" platyrin and its dication. (See Table 2) Both the free
base and dication showed Soret-like bands at U77 nm with very
intense absorptions. They also possessed a much weaker
series of bands in the visible region that were very similar
to the spectra reported for the free base and dication forms

“2 Solutions

of octamethylporphyrin and decamethylsapphyrin.
of 33 and its dication were generally green and dark yellow,
respectively, but they were very solvent dependent. Only on
atnr—visible spectrophotometer can the spectrum of a new
sample be compared with authentic samples in solution. In
each case the same bands were present, but some were a little
sharper, while others might be broader. This was the cause
of the apparent visual changes. Reasonably reproducible
molar absorptivities were obtained for the Soret band, but
it is probably wiser to consider the absorptivities to be
minimum values.

Each time a sample of 33 was dissolved in a solvent or

used for a reaction, a small fraction of the compound would

disappear. The samples did not appear to be light or

36

Table 2. Electronic Absorption of Bis(trimethylene)octamethyl—
platyrin and Related Compounds. 2

 

 

 

Compound Amax' nm (e)
Bis(trimethylene)octa—a u53sh(6o,uoo),A77(398,000),
methylplatyrin 607(11,800), 6u9(9,3u0), 7&7

(2,100), 767sh(1,520), 8A6

(1,850).
Bis(trimethylene)octa-a us3sh(53,200), A77(398,000),
methylplatyrin bis- 625sh(9,030), 637(10,700),
trifluoroacetate salt 6A7sh(9,130), 672(5,190), 688

(6,6A0), 705(7,990), 717sh
(6,220), 73ush(3,630), 788

 

(6,220).
Octamethylporphyrinb 377Sh(55,000), 398(168,000),
”97(1“,U00), 533(loa320), 567
(6,930), 619(5,250)-
Octamethylporphyrinb 3805h(55,000), Ulu(266,700),
dihydrochloride 529sh(3,2U0), 551(16,300), 556
(15,500), 575(6,U90), 597(8,720).
Decamethylsapphyrinb “55(329, 000), 533, 59”, 6H3,
66OSh, 720, 730
Decamethylsapphyrinb N31(56,000), U56.5(59M,000),
dihydrochloride 579(3,UOO), 625(1U,OOO),

677(21,000), 689(17,200).

 

 

aMeasured in methylene chloride.

bMeasured in chloroform.

37

oxygen—sensitive, but weights and spectral values were hard
to duplicate. Solutions of 33 were not durable, and puri-

fied, stabilized solvents were required to minimize losses.
Reagent samples of chloroform are stabilized with 0.7%

“3 while spectro—

ethanol to prevent the formation of phosgene,
grade chloroform and CDCl3 are highly purified, but not
stabilized, solvents. It was believed that early purifica-
tion methods and spectral data were invalidated by trace
quantities of phosgene in our solvents. The main support

for this conclusion was a mass spectrum of 33 that had Just
been removed from chloroform. It gave a peak at m/e 580.
This corresponded to loss of two hydrogens and the addition
of carbon monoxide to the parent peak value of m/e 55A for
.1:-

The mass spectrum of 33 was partially helpful in de-
termining its structure. Around the parent peak value of m/e
554, a large number of large peaks were observed. Hydrogens,
very easily lost from the ring, satisfied this observation,
but made it very difficult to locate the parent peak.

One goal after making a platyrin was to determine its
ability to coordinate with metals. At first glance, the
platyrin should be able to easily incorporate even some very
large metals. However, this did not take into account the
presence inside the ring of two hydrogens attached to the
carbons. (See Figure 3) These hydrogens take up a large

amount of space in the cavity and hinder replacement of

38

 

Figure 3. Internal structure of Bis(trimethylene)octamethyl-
platyrin dication.

the nitrogen protons with metals. Attempts to place copper
(II) and nickel (II) in the ring showed a rapid change in
the color of their solutions. In each case, the uv-visible
spectra were simplified to leave only one or two peaks.

This indicated a loss of aromaticity. Reaction with nickel
(II) acetate gave a band at 522 nm with shoulders at 500

and 5A0 nm. Reaction with copper (II) acetate gave a single
band at 5A8 nm. Neither metal gave stable samples that
could be analysed by IR, PMR or mass spectrometry. Hydrolysis
of these metal complexes with weak to strong acids did not
regenerate the diprotonated platyrin. Each complex gave
similar uv-visible spectra on acid treatment that were not

characteristic of the platyrin. It was possible that the

39

metals sat on the top of the macrocycle and wrapped the
platyrin around themselves like a blanket. If so, the
conjugation of the platyrin would be disrupted, so that the
uv-visible spectrum would show only fragments of the macro—
cycle. Depending on which way the wrapping occurred could
lead to spectra of dipyrrylmethenes (similar to the nickel
complex) or dipyrryltrimethines (similar to the copper com-
plex). This explanation is unlikely, since degradation
of these complexes should have led to completely different
uv—visible spectra for each metal. Yet, the metals did
form complexes. It is conceivable that the complexes were
warped. This could leave one of the meso-bridges suscept-
ible to acid attack. Instead of displacing the metal, the
meso—bridge would be broken and a new product formed. It
is unknown what type of complexes were formed and how stable
they might be to further reactions.

A three-dimensional model was built of the platyrin 5.
It was determined from this model that the two internal
hydrogens on the carbons were situated such that their
electron clouds were almost touching. When the hydrogens
were removed from the nitrogens, the remaining cavities
left room for only a very small atom such as boron. Also,
the configuration of the nitrogen and intruding hydrogens
would allow an atom that coordinated in a tetrahedral shape
the best chance to fit in the macrocycle. Treibs has re-

1
ported2H that dipyrrylmethenes have reacted with boron

U0

trifluoride etherate to give solids with a structure like
3%. The boron was in a tetrahedral configuration and the
three—dimensional model of platyrin suggested that there was

enough room to fit two boron difluoride groups in the two

\ \
N
\B/
I:2
u,

\

N-

cavities, cum: group for each cavity in the interior of the
ring. When the reaction was attempted, a vivid color change
occurred with the growth of two strong bands at 516 and 538
nm. An oily sample was collected by liquid chromatography,
but no Spectral datavnnwaobtained that would elucidate the
structure of the product.

The platyrin can be viewed as an aromatic macrocycle
that may be easily decomposed by several mild reactions.
Its ability to possess a ring current at room temperature
is an unusual, but not unknown, development which may prove
useful to the study of large, planar molecules. The lack of
stability may hinder its use as a ligand. While, its simil—
arity to the porphyrins is now confined to PMR and uv-visible
spectra, possibly, a large supply of a platyrin would enable

the study of its reactions to electrophilic substitution.

‘3" __

Al

The overall goal of achieving a 22 w—electron system capable
of aromaticity at room temperature was accomplished with
this synthesis. The use of the porphyrin system as a model
' proved successful and it may be a critical stepping-stone to

the determination of aromatic character in large macrocycles.

EXPERIMENTAL

General Procedure

 

The melting points were determined on a Thomas Hoover
Unimelt melting point apparatus and are uncorrected.

The infrared spectra were recorded on a Perkin-Elmer
Model 237B spectrophotometer. The PMR spectra were obtained
on Varian T—6O and Brucker 180 spectrometers with chemical

shifts reported in 6-units measured from tetramethylsilane

 

as the internal standard. The CMR spectra were obtained on
a Varian CFT—2O spectrometer with chemical shifts reported
in S—units measured from tetramethylsilane as the internal
standard. The UV-Visible spectra were recorded using a
Unicam Model SP—800 spectrophotometer using 1/2 cm quartz
cells, or using a Cary 17 spectrophotometer using 1 cm
quartz cells. Mass spectra were obtained with a Hitachi
Perkin—Elmer RMU-6 mass spectrometer. Microanalyses were
not obtained due to the extreme unstability of the com—

pounds.

2,2'—Dipyrrylmethane (p)

 

The procedure of Clezy30’31

was used and the product
was purified by crystallization from petroleum ether (30-
60°C), mp 72—3°C. Compound 6 was air-sensitive and stored

at -20°C.

H2

“3

5,5'-Bis(3eoxo—butenyl)-2.2'—dipyrrylmethane (1%)}?

Compound 6 (0.500 g) and 3-butynone (0.700 g) were dis—
solved in 125 ml of oxygen—purged methanol and refluxed
under nitrogen for 68 hours. Water was added and the yellow
precipitate collected by filtration. Recrystallization
from ethanol-water gave 0.3035 g (31.H% yield) of 7a: mp
210°C (dec.); IR (KBr): 3265 cm'1(N-H), 1665 cm'l(c=0),
1615, 1555, and 1u75 cm'1(c=c); PMR (DMSO-d6): 67.15 (d,
J = 15 Hz, 2H, trans HgC=C-H), 6.28 (m, 2H, B-pyrrolic hydro-
gens), 6.22 (d, J = 15 Hz, 2H, trans 57C=C-H), 5.80 (m, 2H,
B—pyrrolic hydrogens), 3.83 (s, 2H, -C§2—), 2.15 (s, 6H,

CH —C=0); UV—Vis Amax (CH2C12): 350 nm; mass spectrum (70

3
eV): m/e 282 (parent).

5,5'—Bis(3-oxoepropenyl)-2,2'-dipyrrylmethane (7p)32

Compound 6 (0.10U5 g) was dissolved in 2.5 ml of oxygen-
purged methanol under nitrogen. A solution of 0.0847 g of
propargyl aldehyde33 in 2.5 ml of oxygen—purged methanol
was slowly dripped into the rapidly stirred solution of 6.
The reaction was stirred at room temperature for U8 hours.
Then 10 ml of water was added and the precipitate collected
by filtration. The powder was chromatographed on silicic
acid with methylene chloride—ethyl acetate (1:1) as the
eluant and recrystallized from ethanol-water to give 51.6

mg (28.U% yield) of 1p: mp 203°C (dec.); IR (Nujol): 3210

uu

cm‘1(N-H), 1655 cm'1(c=0), 1620, 1590, and 1555 cm‘1(c=0);

PMR (DMSO-d6): 09.23 (d, J = 8 HZ, 2H, -CflQ), 7.27 (d, J =

15 Hz, 2H, trans H70=C—H), 6.U5 (m, 2H, B-pyrrolic hydrogens),
6.32 (0, 2H, trans HgC=C-H), 5.88 (m, 2H, B—pyrrolic hydrogens),

3.90 (s, 2H, —032-); UV—Vis xma (CH2012): 352 nm; mass

X

spectrum (70 eV): m/e 25H (parent).

Attempted Synthesis of 2,2'-Dipyrrylpropane (8)

 

Stoichiometric amounts of pyrrole and 1,3,3-trimethoxy-
propene were dissolved in acetic acid and let sit until deep
blue (Amax 580 nm). A deep blue tar was collected by addi-
tion of water, but its reduction to 8 was not attempted.

The same results were found upon reaction of 6 and 1,3,3-

trimethoxypropene.

Triphenylmethylfluoroborate (12)

 

The procedure of Dauben, gt al.u5 was used to obtain

the orange solid, 12. The product was stored in a desiccator.

2,5-Bis(triphenylmethyl)pyrrole (l3)

 

Compound 12 (6.60 g) was dissolved in 200 ml of aceto—
nitrile. Pyrrole (0.70 g) was added dropwise to the stirred
solution at room temperature. Colorless crystals formed im—
mediately. The solution was cooled to 0°C in an ice-water

bath and the crystals were collected by filtration.

45

Recrystallization from chloroform-hexane gave 5.50 g (99.8%
yield) of i3: mp 229-30°C; IR (Nujol): 3430 cm‘1(N-H),
1600 and 1540 cm—l(C=C), 1080 and 1035 cm"1(mono-substituted
benzene); PMR (CDCl3): 67.05 (s, 30H, phenyl), 5.78 (d, J =
2 Hz, 2H, B-pyrrolic hydrogens); CMR (CDC13): 6145.9, 137.2
(a-pyrrolic carbon), 130.2, 127.6, 126.3, 109.2 (d, B-pyr-
rolic carbon, 60.4 (tetra—substituted methyl); mass spectrum

(70 eV): m/e 551 (parent).

Dimethyl—N,N'-bis(tosylmethyl)glutarimidate (léalig

Compound 18 (2.20 g) was dissolved in 30 ml of nitro-
methane under nitrogen with slight warming. Methylfluoro-
sulfonate (4.39 g) was quickly added and the reaction stirred
for 16 hours. The solvent was removed under reduced pres—
sure along with excess methylfluorosulfonate. The residual
oil was dissolved in CHC13. This solution was washed with
very dilute HCl (20:1), dried with MgSOu and the solvent
was removed. The remaining oil gradually turned waxy with
sitting. The product obtained was the bisfluorosulfonate
of 16%: PMR (DMSO—d6): 58.62 (t, J = 7 Hz, 2H, HN-CH2),
4.57 (d, J = 7 Hz, 4H, HN-CHZ), 3.93 and 3.88 (6H, syn-

and anti—OCH3).

Dimethyl—N,N'-bis(tosylmethy1)thioglutarimidate (163135

The same procedure was used as for 16a, using 19 as
’V‘UL N

starting material, but no product was isolated.

46

N,N'-Bis(tosylmethyl)glutaramide (l8)

 

Sodium p—toluene sulfinate (80.0 g), 37% formaldehyde
(35 m1), and glutaramide (26.0 g) were mixed together in 400
ml of water and heated to 90°C. To this hot solution was
added 100 ml of 90% formic acid. After 16 hours the reaction
was cooled to 0°C and the crystals were collected. Recrystal-
lization from CHCl3 gave 43.8 g (47% yield) of 18: mp 156-7°C;

l l 1

IR (Nujol): 3300 cm” (N-H), 1655 cm” (C=0), 1535 cm"

(R—NH¢C=0); PMR (DMSO-d6): 68.70 (t, J = 6 Hz, 2H, N-H),

4.58 (d, J = 6 Hz, 4H, CH —NH), 2.35 (s, 6H, CH3—), 1.97 (t,

2
J = 6 Hz, 4H, CH2-CH2-C=O), 1.50 (m, 2H, CH2-CH2-C=0).

NJN'-Bis(tosylmethyl)thioglutaramide (%%)

 

Compound 88 (25.00 g) was added to 700 ml of dioxane
and heated to 60°C with stirring under nitrogen. Phosphorus
pentasulfide (24.00 g) was added and the solution was stirred
for 7 hours. The orange solution was filtered and the sol—
vent removed under reduced pressure. The residue was re—
crystallized from chloroform-ethanol to give 24.40 g (98.7%
yield) of a pale yellow solid, 8%: mp 154—5°C; IR (Nujol):
3260 cm-1 (N—H), 1510 cm-1 (mono—substituted benzene), 1270

cmfl (0:3); PMR (DMSO-d6): 510.50 (t, J = 6 Hz, 2H, CH —N-H),

2

-CH _

5.23 (d, J = 6 Hz, 4H, CHZ-NH), 2.43 (t, J = 7 Hz, CH 2

2
C=0), 1.73 (p, J = 7 Hz, 2H, CH2-CH2—C=O).

47

N,N'—Bis(glycyl)glutaramide (8%)

Glycine (7.5 g) was dissolved in 40 m1 of 2.5 N NaOH
and cooled with an ice bath to 0°C. In separate addition.
funnels were placed 8.45 g of glutaroyl chloride and 40 ml
of 2.5 N NaOH. Both were dripped into the stirred solution
at the same rate over 5 minutes. The solution was stirred
an additional 10 minutes, then acidified to Congo Red end
point with conc. HCl. The solution was cooled to 0°C and
filtered to give 8.7 g (71% yield) of white crystals, 8%:
mp 172-4°C; PMR (DMSO-dé): 68.00 (t, J = 6 Hz, 2H, CHz-NH),

3.70 (d, J = 6 Hz, 4H, CHZ—NH), 2.13 (t, J = 6 Hz, 4H, CH2

CH2—C=O), 1.82 (p, J = 6 Hz, 2H, CH2-CH —C=O).

2

Attempted Synthesis of 1,3-Bis(3,4-dicarbethoxypyrrol-2-yl)-

propane (8%)37

Compound 8% (1.23 g) and dimethylacetylenecarboxylate
(1.25 ml) were mixed together in 15 m1 of acetic anhydride
and heated at a variety of temperatures from 60°C to reflux.
When the solvent was removed under reduced pressure, the re-

maining oil could not be identified.

3,4-Dimethylpyrrole (88)

 

The procedure of IchimurauO was followed except for
the following changes: 1) When the pyridine and thionyl

chloride were added to the reaction, the thionyl chloride

 

48

was added slightly ahead of the pyridine. 2) The oily resi—
due of 2—ethoxycarbonyl-3,6—dihydro-4,5-dimethyl—l,2-thiazine—
l—oxide was added to 135 g of potassium hydroxide dissolved

in 300 ml of methanol. If the second step was not followed,
the reaction would explode through the condenser. Typical
yields were 50-55% based on ethylcarbamate. The product was
vacuum~distilled with an air condenser, since it melts at
slightly above room temperature. Compound 88 may be stored

at least 6 months at 0°C under nitrogen without discoloring.

3,4,3',4'-Tetramethyldipyrryl-2,2'-hexacyclotrimethine

 

tetrafluoroboratengé)27

 

l,3-Cyclohexanedione (2.75 g) was dissolved in 20 ml
of ethanol and filtered to remove the stabilizing reagent.
Compound 88 (4.50 g) was dissolved in 80 m1 of ethanol and
the two solutions were mixed together and heated to reflux.
Tetrafluoroboric acid (8 ml) was added dropwise to the re~
action and the heating continued for 5 minutes. The reaction
was cooled slowly. It was allowed to stand at room tempera-
ture for 4 days protected from light. The deep blue crystals
were collected by filtration and washed repeatedly with ethyl
ether until the ether washings were very pale blue. There
was obtained 2.9 g (35% yield) of 88: mp 245-6°C; IR (Nujol):
3350 cm”1 (N—H), 1565 and 1535 cm‘1 (C=C); PMR (DMSO-d6):
67.5 (d, J = 3 Hz, 2H, a—pyrrolic hydrogens), 7.18 (s, 1H,
C=CH—C); 2.9 (t, J = 5 Hz, 4H, CH2-C=C), 2.32 (s, 6H, 4,4'-

49

CH3), 2.00 (s, 6H, 3,3'-CH3), 1.9 (m, 2H, cg2); CMR (DMSO-d6):
5158.9, 135.5, 134.7, 132.2, 126.2, 116.6 (a-pyrrolic carbon),
27.4, 21.0, 18.6, 13.8, 9.8; UV-Vis Amax (CHC13): 540sh and
580 nm; mass spectrum (70 eV): m/e 266 (parent — HBFu).

The method given was a general procedure for the prepara—
tion of dipyrryltrimethines. The same ratio of reactants was
used to prepare 3,5,3',5'—tetramethyl-4,4'-diethy1dipyrryl-
2,2'-hexacyclotrimethine tetrafluoroborate (38) from 2,4-
dimethyl—3—ethylpyrrole and l,3-cyclohexanedione in 63% yield:
mp 213-3°C; IR (NuJol): 3350 cm‘1 (N-H), 1535 and 1520 cm'1

(C=C); PMR (CDC13): 57.08 (s, 1H, C=CfleC), 2.88 (t, J = 6 Hz,

4H, CH -C=C), 2.43 (s, 6H, 5,5'-CH3), 2.4 (q, J 8 Hz, 4H,

2
CH3-CH2—), 2.30 (s, 6H, 3,3'-CH3), 1.03 (t, J = 8 Hz, 6H,

cg -CH2-); CMR (CDC13): 5154.4, 146.2, 135.0, 131.1, 129.8,

3

112.9, 27.5, 22.0, 16.9, 14.6, 13.7, 11.8; UV—Vis ax (CHC13):

Am
5603h and 599 nm; mass spectrum (70 eV): m/e 322 (parent -

HBFM).

 

1,3-Bis(3,4-dimethy1pyrrol-2-y1)cyclohexane (81)

Compound 88 (2.5 g) was mixed with 5 g of sodium acetate
and 0.1 g Pd/C. Ethanol (100 ml) was added and the mixture
was shaken on a Parr Hydrogenator at 50 psi for 5 hours. The
mixture was filtered through anhydrous K2003 and the solvent
removed under reduced pressure to give a pale yellow oil of
€1= PMR (CDC13): 67.27 (broad, 2H, N-H), 6.20 (0, 2H, a—pyr-
rolic hydrogens), 1.97 (s, 6H, 4,4'—CH3), 1.92 (s, 6H, 3,3'-

50
0&3), 2.0-1.0 (corresponded to a 1,3-dia1kylcyclohexane).
Compound 31 was always used immediately since it was

air—sensitive.

l,3—Bis(5-formyl-3,4-dimethylpyrrol-2eyl)cyplohexane (g8)

 

Calcium hydride-dried dimethylformamide (1.7 ml) was
cooled to -5°C under nitrogen with stirring. Phosphorus
oxychloride (2.0 ml) was added dropwise by syringe. The mix-
ture generally solidified and was dissolved in 10 m1 of di—
chloroethane. A sample of £1 was prepared as reported above
and dissolved in 10 m1 of dichloroethane. This solution was
added dropwise to the cold reaction mixture. After stirring
at 0°C for 15 minutes it was then refluxed for 15 minutes.
After slight cooling the reaction mixture was added to a solu-
tion of 35 g of sodium acetate in 200 m1 of water, which was
heated at reflux for 15 minutes. The resulting oil was ex-
tracted 3 times with chloroform. The extracts were combined
and washed twice with a saturated sodium chloride solution.
The extracts were dried with anhydrous K2C03 and the solvent
removed under reduced pressure. The oil was crystallized
from chloroform-pentane to give 1.86 g (80.6% yield) of a
yellow solid, 88: mp 220-2°C; IR (NuJol): 3260 and 3210
cm‘1 (N-H), 1640 cm"1 (C=0); PMR (CDC13): 59.33 (s, 2H, -CH0),
2.87 (m, 2H, tertiary CH), 2.20 (s, 6H, 4,4'-CH3), 1.90 (s,
6H, 3,3'—CH3), 1.87 (m, 8H, l,3-disubstituted cyclohexane);

UV-Vis ) (CHC13): 313 nm; mass spectrum (70 eV): m/e

max

 

 

51
326 (parent).

3,4,8,9,17,18,22,23:00tamethy1-30,31,33,34-tetraagaheptacyclo-
[23.3.1.12’5.17’10.111’15.116’19.121’2u1tetratriaconta-

2,4,648,10(33),16,18,29,22,24(3O)-decaene(88)

The actual preparation involved the synthesis of the
bis-tetrafluoroborate salt which gave g2 on deprotonation.

Compound 81 (0.38 g)(prepared from 0.50 g of 88 as
described above) and Compound 88 (0.462 g) were dissolved
in 200 ml of acetic acid. The solution was heated at reflux
for two hours with air bubbling into the reaction. Then
1 ml of 48% HBFu was added and the reflux continued for 16
hours. The reaction was cooled to room temperature and let
sit for 24 hours. The metallic blue (overtones) crystals
were collected by filtration to give 0.2468 g (23.7% yield)
of the HBFu salt of 82. The salt may be neutralized by dis-
solving the blue crystals in ethanol and treating with aque-
ous sodium hydroxide followed by extraction with chloroform.
After removal of the solvent, the yellow solid was chroma-
tographed on basic alumina with chloroform and the first

1 (N-H),

band was 82: mp 301-4°C; IR (NuJol): 3350 cm-
1620 cm’1(C=C), PMR (CD013): 56.50 (s, 2H, meso—methene),
2.07 (s, 12H, cg3—), 1.92 (s, 12H, CH3—), 3.0-1.0 (c, 20H,

1,3-disubstituted cyclohexanes); UV—Vis A (CHC13): 429

max

nm for 82, 462 nm for HBF4 salt; mass spectrum (70 eV): m/e

560 (parent).

 

52

l,3-Bis(3,4—dimethy1pyrrol—2-yl)cyclohex—l-ene (3Q)ul

 

Compound g8 (2.50 g) and 2.5 g of sodium borohydride
were mixed together in 200 ml of acetonitrile. The mixture
was heated on a steam bath until all the blue color was
gone (girga 5 minutes). The solvent was removed under re-
duced pressure and the residue was washed with an ether—water
mixture. The ether solution was dried with anhydrous potas—
sium carbonate and filtered. The solvent was removed under
reduced pressure to give an oil of essentially quantitative
yield, 88. The oil was pale yellow, but quickly became green
on exposure to air: IR (neat): 3380 cm_1 (N—H), 1625 and 1580
cm“1 (C=C); PMR (00013): 57.4 (broad, 2H, N—H), 6.25 (c, 2H,
a—pyrrolic hydrogens), 5.67 (c, 1H, -C=C-H), 3.60 (c, 1H,
CHZ-C=C), 2.4-1.5 (6H, l,3—disubstituted cyclohexene), 2.07
(s, 3H, Cg3- closest to vinyl), 1.98 (s, 9H, CH3-).

Compound 88 was never stored but immediately used to

make 88 or 88.

1,3-Bis(5—formy1—3,4-dimethylpyrrol-2-yl)cyclohgx—l-ene (3g)

From a sample of 88, as prepared above, was made 8%, in
the same manner as for the synthesis of 88’ Compound 88
was made in 80.2% (1.84 g) yield, based on 2.50 g of 38,
through this procedure: mp 174—6°C (dec.); IR (Nujol):

3250 cm‘1 (N-H), 1630 cm‘1

(C=O); PMR (DMSO—d6): 59.40 (s,
1H, —CHC), 9.35 (s, 1H, -CHO), 5.83 (m, 1H, C=C—H), 3.6

(m, 1H, Gag—C=0), 2.22 (s, 6H, 4,4'—cg3), 1.98 (s, 3H, 3-033),

53

1.93 (s, 3H, 3'-Cfl3), 2.6-1.6 (0, 6H, 1,3-disubstituted cyclo—

hexene); mass spectrum (70 eV): m/e 326 (parent).

3,4,8,9,17,18,22A23-Octamethyl-30,31,33,34-tetraazahepta-
11,15 16,19 21,24

 

cyclo[23.3.1.12’5.17’10.l 1 1 Jtetra-

 

triaconta-1,3,5,7(33),8,10,15(32),l6,l8,20,22,24(30),
25(29)-tridecaene (3%)

 

 

Compound 88 (0.1092 g) was reduced, as shown previously,
with NaBHu to give 88 which was dissolved in 20 ml of methanol.
Compound 8% (0.0992 g) was dissolved in 20 ml of dichloro-
methane and diluted with 40 ml of methanol. The two solu-
tions were mixed together and added dropwise to a refluxing
solution of 5 m1 of 48% HBr in 100 m1 of methanol with air
bubbling into the solution. After the addition was complete,
another 2 m1 of HBr was added with refluxing and air bubbling
continued for 30 minutes. The reaction then sat in the dark
for 36 hours. The product was collected and neutralized by
diluting the reaction with 200 m1 of water and 20 ml of 30%
NaOH. The solution was extracted 2 times with dichloromethane.
The CH2C12 layer was condensed and chromatographed on neutral
alumina ("Baker Analyzed" Reagent) with dichloromethane,
followed by 1% methanol in dichloromethane. The green band
was collected and the solvent removed to give 32.2 mg (19%

yield) of £3: mp >300°C; PMR (CDCl —CF3C02H): 611.64 (s,

3
2H, meso-Cfl), 4.22 (s, 12H, Cfl3-), 4.17 (s, 12H, Cfl3-),

2.4-1.0 (c, 12H, corresponded to 6 C32), -5.64 (s, 4H, N-H),G

54

-8.97 (S, 2H, internal C=Cfl3C); UV-Vis Am x (CH2C12): ,230

a
nm (8 13,600), 265 (12,000), 276 (11,900), 298 (11,400), 326
(11,900), 3905h (1“,200): “53 (603u00)9 ”77 (398,000): 607
(11,800), 649 (9340), 747 (2100), 767sh (1520), 846 (1850):

UV—Vis Ama (CH2C12)(bis-tetrafluoroborate salt): 231 nm

x
(8 19,900), 281 (9130), 290 (8920), 307 (7470), 375 (10,800):
408 (10,500), 453sh (63,200), 477 (398,000), 625sh (9030),
637 (10,700), 647sh (9130), 672 (5190), 688 (6640), 705
(7990), 717sh (6270), 734sh (3630), 788 (6220); mass spectrum
(70 eV): m/e 552 (parent — 2).

When dissolved in DMSO-d6 with a D 0 contamination, the

2
PMR signals at 611.64, -5.64, and -8.97 disappeared. The

signals reappeared after a H20 wash.

Attempted Complexation of Compound 8%

 

Samples of 88 were treated with nickel (II) acetate or
copper (II) acetate in dimethylformamide at reflux. In each
case, a color change was immediately visible. Solid samples
were collected by addition of water followed by filtration.
No chemical characteristics could be determined for either
sample due to a lack of solubility in solvents such as CHC13,
CH2012, or benzene. Addition of acetic, trifluoroacetic or
sulfuric acid to either complex did not regenerate Compound
53,-

Nickel(II)-platyrin complex: UV-Vis Amax (DMF): 522 nm.

Addition of acid gave a peak at 490 nm.

55

Copper(II)-p1atyrin complex: UV—Vis Am (DMF): 548

ax
nm. Addition of acid gave a peak at 490 nm.

Reaction Between Compound ééand Borontrifluorideuu

 

 

A sample of 33 (10 mg) was dissolved in dry benzene (20
ml) and added to a solution of triethylamine (2.0 ml) and
borontrifluoride etherate (1.8 m1) under nitrogen. The re-
action was heated for 10 minutes at reflux, then cooled to
room temperature and washed 3 times with water. The benzene
layer was dried with MgSOu and the solvent removed under re—
duced pressure to leave a red oil. The oil was chromatographed
on neutral alumina with methylene chloride, followed by 2%
methanol in methylene chloride. The major red band was col-
lected and the solvent removed to give a red tar: UV-Vis
Amax (CH2012): 516 and 538 nm. No further data could be
obtained by IR or PMR spectrometry due to the small amount

of material.

BIBLIOGRAPHY

10.

11.

12.

13.
14.

15.

16.

17.
18.

19.

BIBLIOGRAPHY

A. Kekule, Bull. Soc. Chim. France, 8, 98 (1865).

 

. A. Kekule, Ann., 888, 77 (1872).

Willstatter and M. Heidelberger, Ber., 88, 517 (1913).

. W. Armit and R. Robinson, J. Chem. 800., 12Z, 1604
1925). mm

H. Heitler and F. London, Z. Phys., 88, 455 (1927).

 

F
F
J. Thiele, Ana); 888, 87 (1899).
R
J
(

E. Huckel, z. Physik., 18’ 204 (1931); 72, 310 (1931).

 

D. P. Craig, Non-Benzenoid Aromatic Compounds, D. Ginsburg,
ed.,u(New York: ’Interscience Publishers, Inc., 1959),
p. 1 .

 

L. Pauling, Perspectives in Or anic Chemistry, A. R. Todd,
ed. (New York: InterscIence uBlishers, Ific., 1956),

 

 

’
6
. J. s. Dewar and G. J. Gleicher, J.A.C.S., 81, 685 (1965);
H Chung and M. J. S. Dewar, J. Cfiem. PEys., 88, 756
6 .

C. C. Leznoff and F. Sondheimer, J.A.C.S., 82, 4247
(1967).

R. M. McQuilkin, B. W. Metcalf, and F. Sondheimer, Chem.
Comm., 821%, 338.

M. J. s. Dewar, J.A.C.S., 13, 3353 (1952).

 

R. Breslow, J. Brown, and J. J. Gajewski, J.A.C.S., 82,
4383 (1967). ———-————

M. J. S. Dewar, Aromaticity Special Publication No. 21
of the Chemical SBCIety, London, pp. 177-198 (1967).

 

H. P. Figeys, Tet., 88, 5225 (1970).

L. Pauling, J. Chem. Phys., 8, 673 (1936).

 

F. London, J. Phys. Radium., 8, 397 (1937).

 

E. D. Bergman, J. Horau, A. Pacault, A. Pullman, and B.
Pullman, J. Chim. Phys., 8%, 474 (1952).

 

56

20.
21.

22.

23.

24.
25.

26.

27.
28.

29.

30.

31.

32.
33.

34.
35.

36.

37.

38.

39.

57

A. Pople, J. Chem. Phys), 58, 1111 (1956).

 

Sondheimer, Acct. Chem. Res., 8, 81 (1972).

 

J
F
L. M. Jackman, F. Sondheimer, Y. Amiel, D. A. Ben-Efraim,
Y. Gaoni, R. Wolovsky, and A. A. Bothner-By, J.A.C.S.,

85, 4307 (1962). '

J. Griffiths, Colour and Constitution of Organic Molecules
(New York: Academic Press, Inc., 1976), p, 226;

 

 

I. C. Calder and F. Sondheimer, Chem. Comm., 8288, 904.

R. M. McQuilkin, P. J. Garrett, and F. Sondheimer, J.A.C.S.,
2%, 6682 (1970).

B. W. Metcalf and F. Sondheimer, J.A.C.S., 2%, 5271 (1971).
A. Treibs and E. Hermann, Ann., 882, 207 (1954).

M. Strell and F. Kreis, Chem. Ber., 81, 1011 (1954).

 

K. J. Reed, U.S. Patent 2,739,148; Chem. Abstr., 8%, 901c
(1957).

P. S. Clezy and G. A. Smythe, Aust. J. Chem., 22, 239
(1969). ”“

R. Chong, P. S. Clezy, A. J. Liepa, and A. W. Nichol,
Aust. J. Chem., 88, 229 (1969).

 

 

 

 

A. W. Johnson, J. Chem. Soc., 8281, 1626.

J. C. Sauer, Org. Syntheses Coll. Vol. 4, N. Rabjohn, ed.,
(New York: John Wiley and Sons, Inc., 1963), p. 813.

 

E. Khotinsky and R. Patzewitch, Ber., 88, 3104 (1909).

A. NL ‘van Leusen, H. Siderius, B. E. Hoogenboom, and D.
van Leusen, Tet. Let., 8218, 5337.

 

H. A. Hovwing, J. W. Idman, and A. M. van Leusen, Tet.
Let., 8218, 143.

R. Hulagen, H. Gotthardt, H. O. Bayer, and F. C. Schafer,
Chem. Ber., 88%, 2611 (1970).

 

H. O. Bayer, H. Gotthardt, and R. Hulagen, Chem. Ber.,
$93, 2356 (1970).

F. M. Hershenson, J. Org. Chem., 88, 1260 (1975).

 

 

40.

41.

42.

43.

44.
45.

58

K. Ichimura, S. Ichikawa, and K. Imamura, Bull. Chem. Soc.
Japan, 82, 1157 (1976).

C. Jutz, A. F. Kisschner, and R. M. Wagner, Chem. Ber.,
11g, 1259 (1977).

M. M. King, Ph.D. Thesis, Harvard University, 1970.

 

L. F. Fieser and M. Fieser, Reagents for Organic S nthesis,
Vol. 1, (New York: John Wiley and Sons, Inc., 19 7 ,
p. 130.

A. Treibs and F. H. Kreuzer, Ann., 118, 208 (1968).

H. J. Dauben, Jr., L. R. Honnen, and K. M. Harmon, J. Org.
Chem., 88, 1442 (1960).

 

59

2.5 3.0 3.5 4.0 MICRONS 5.0 5.0 8.0
100 ' ' ' '

 

 

       

80
.3
E’
U50
2 ..
<
1...-
:
5.
240
<
a:
p—
20
0 1 : i : ; . ‘ . ' i i. ' 3 ’ '3 i,
4000 3500 2500 2000 1500
FREQUENCY (CM ')

5.0 5.0 7.0 8.0 M'CRONS 10.0 11.0120 15.0
100 -- . . -; “. , 100
80 80

5‘

E'

U

250 50
< .-

.—

t

5.

z

<40 40
M

p.-

20 20

0 j , _ . . , . . . . . o
2000 1800 1500 1400 1200 1000 800

"EOUENC Y 'CM '1

Figure 4. Infrared spectrum of 5,5'-Bis(3-oxo-buteny1)-2,2'-
dipyrrylmethane (Z8)'

60

2.5 3.0 3.5 4.0 M'CRONS 5.0 5.0 8.0
100

TRANSMITTANCE(°/o)

 

o 7 ’ ' l . ' I ‘ ' ' . 3 ’ 2 T ’ . ’f
4000 3500 3000 2500 2000 'l 500

neoumcv (CM '1

5.0 5.0 7.0 8.0 M'CRONS 10.0 11012.0 15.0
. . . ‘00

 

 

80 80
33.
LU
U
2 60 60
<
p—
:2
E
‘2 a
< 40 40
M -
y-

20

N
O

  

0 1ézzt:-:-_:-.1
2000 1800 1600 1400 1200 1000 800

FIEOUENCV (CM ')

 

Figure 5. Infrared spectrum of 5,5'—Bis(3—oxo—Dropenyl)-2,2'-
dipyrrylmethane (10).

61

2.5 3.0 ’ 3.5 4.0 M'CFONS 5.0 5.0 8.0
100 ‘ ' 100

©

8
0 » .

20

N
O

80 80
§
I.“
U60 60
Z
< _.
p.-
:
2
m _
E40 . © \ 40
as .©° (N) °‘©

g H

- 4
. I
4
l
| .
u
u I
. A

 

I

 

0 : . . . . . . . . : 2 . . . .. . . 0

4000 3500 3000 2500 2000 1500
FREQUENCY (CM '1

5.0 5.0 7.0 8.0 M'CRONS 10.0 11012.0 15.0

100

80 80
§
3'
U
2 60 60
<
.—
’2
ii
‘2
< 40 40
M
p—

20

N
O

 

o. . . ' 1 - : 2 . ' i E - ; .. t - 0
2000 1800 1600 1400 1200 1000 800

ntoumcv «CM '1

Figure 6. Infrared spectrum of 2,5-Bis(triphenylmethyl)—
pyrrole (IQ).

62

2.5 3.0 3.5 4.0 M'CRONS 5.0 5.0 8.0
100 ' '

 

 

 

80
83
if
U 50
Z
<
.—
:
3
z 40
<
a:
.—
20
4000 3500 3000 2500 2000 1500
FIIOUIKY (CM')
5.0 5.0 7.0 8.0 WCWNS 10.0 11.0 12.0 15.0
100 ' ‘ 100
‘ 9
Tomi-1211140cuZCHZCHZCNHCHZTos
80 _ ' '

3

Lu

U

2 60 60

<

.—

t

z

‘2 ,

< ‘0 40

3

.—

20

 

2000 1800 1600 1400 1200 I“ m

"(coma (CM '1

Figure 7. Infrared spectrum of N,N'-Bis(tosylmethyl)-
glutaramide (IQ).

63

2.5 3.0 3.5 4.0 M'CRONS 5.0 5.0 8.0

100 ' z . . , , ' 100
§ 2‘
TOOCHZNHCCHZCHZCHZCNHCHZTOO

80 '.;123=;'=1293 80
§‘
350 50
Z
<
C
.%
240 40
<
a:
.—

20

N
O

 

0 ' . ': 5 ' . .1 _ . . .- , ,
4000 3500 3000 2500 2000 1500
FIEOUINCY (CM ')

5.0 5.0 7.0 8.0 ““9” 10.0 11012.0 15.0

100 ‘ ‘ 100
§ $3 “I
15.0H2uHccuzcuzcHzcuucnzToa

80 80
§‘
I;
U
260 60
<
.—
t
i
3
<40 -. 40
a:
.—

 

0 ' . ‘ . ' 5‘. ’ : , - .
2000 1800 1600 1400 1200 1000 800

"£00!ch (CM '1

Figure 8. Infrared spectrum of N,N'-Bis(tosylmethyl)thio-
glutaramide (IQ).

614

2.5 _ 3.0 3.5 g 4.0 M'CRONS 5.0 5.0 8.0
100 I ‘ I : 1 ' . 5 . . . 1

 

 

80
Es"
E’
U 50
Z
<
)—
:-
5.
z 40
<
K
.—
20
o . i . . . 1 . _ . . u I . . ' . . ' :
4000 3500 3000 2500 2000 1500
moumcv (CM ’1
5.0 5.0 7.0 8.0 M'CRONS 10.0 11012.0 15.0
100 100
9 9
Hozccnznaccuzcazcuzcuncuzcozn
80 _ ' 80
E
m
U
Z50 50
<
.—
L-
2
'2 _
<40 _ 40
K
p—
20 20
o - - . . o . ! . O
2000 1800 1500 1400 1200 1000 800

mouencv (CM '1

Figure 9. Infrared spectrum of N,N'-Bis(glycyl)glutaramide

(€13).

65

2.5 3.0 3.5 4.0 M'CRONS 5.0 5.0 8.0
100 ‘

b 0 (D
O O O

TRANSMITTANCE(%)

N
O

 

0 t ‘ ' ' : f . . T . t ' ' : . ‘ . ' *
4000 3500 3000 2500 2000 I 500
PICOUCNCV (CM ')

5.0 5.0 7.0 8.0 M'CRONS 10.0 11.0120 15.0
* 100

80 . 80

8

TRANSMITTANCE(%)
8

20 9 20
8 F4

I
4

    

O ‘ . . - f I . . . . . . . a
2000 1800 1600 1400 - 1200 1000 800

'IEOUINCV (CM ')

 

FiEure 10. Infrared spectrum of 3,h,3',U'-Tetramethyldipyrryl—
2,2'-hexacyclotrimethine tetrafluoroborate (€Q)-

66

2.5 . 3.0 3.5 4.0 M'CRONS 5.0 5.0 8.0
100 '

8 8 8

TRANSMITTANCE(%)

N
O

 

0 . . . . . . . . . . . . . . ,
4000 3500 3000 2500 2000 1500
IIEOUENCVKM')
5.0 5.0 7.0 8.0 M'CRONS 10.0 11.0120 15.0
100 ' 100

80 80
2';
u;
u
z 50 60
<
.—
t:
:2
‘2
<40 40
8
.-

20 IIH

CHO

 

0:45 f : ' = . : " - ' - z ; 4 - . ~ - :
2000' 1800 1600 1400 1200 1000 800

"(QUINCY (CM ‘1

Figure 11. Infrared spectrum of 1,3—Bis(5—formyl-3,u-dimethyl-
pyrroI-2-yl)cyclohexane (QQ).

67

2.5 3.0 3.5 4.0 MICRONS 5.0 5.0 8.0

0
O

TRANSMITTANCE(%)
3.
o

M
O

 

4000 3500 3000 2500 2000 1 500

HEOUENCVKM‘)
5.0 7 8.0 M'CRONS 10.0 11.0

TRANSMITTANCE(%)

 

1800 I600 1400 1200 1000 800

IIEOUENCV [CM ')

Figure 12. Infrared spectrum of 3,“,8,9,17,18,22,23—Octamethyl-

3O,31,33,3U—tetraazaheptacyclo[23.3.1.12’5.17’1O-

.111’15.116’19.121’ZMJtetratriaconta-2,U,6,8,10—
(33),l6,l8,20,22,2U(30)-decaene (g2).

68

2.5 3.0 3.5 4.0 MICRONS 5.0 5.0 8.0

S

TRANSMITTANCE(%)

 

0
4000 3500 3000 '2500 2000 I 500

'llOlJfNCV (CM ‘1

5.0 7.0 8.0 M'CRONS 10.0 11012.0

TRANSMITTANCE(%)

 

1800 1600 1400 i 200 1000 800

ntootucv (CM ')

Figure 13. Infrared spectrum of 1,3—Bis(5—formyl-3,4—dimethyl—
pyrrol-2—yl)cyclohex—l-ene (ég).

69

2.5 3.0 3.5 4.0 M'CRONS 5.0 5.0 8.0

  

80
o3
E
U 50
Z
<
.—
t
E.
2 40
<
K
.— r

20

0 .

4000 3500 3000 2500 2000 1500

"(QUINCY (CM')
50 10 80 MKRONSIOD 1Lo1zo

 

‘I'RANSMITTANCE(%)

 

1800 1600 1400 i200 1000 800

5.600th (CM ‘)

Figure 1“. Infrared spectrum of 3,5,3',5'—tetramethyl-U,U'—
diethyldipyrryl-2,2'—hexacyclotrimethine tetra—
fluoroborate (%&).

 

 

 

 

 

 

 

WIVVVV' Iv rlv rvIvyv I v lvvrvIvvvv

L IL gm
no

',

NH
NH ;
/’ 2’ i
0

I l 1 l J L
I .l. .41.. .I...LIA. l I 1 LJ_.
no 10 to u ”I” u so 20 lo a

Figure 15. PMR spectrum of 5,5'—Bis(3-oxo—butenyl)-2,2'-
dipyrrylmethane (1Q).

 

vivi‘v—rv‘vrvrvvvvlvvvvlfvrv’Iva‘vIvu .1..‘.I..v.

 

 

 

 

 

   

 

. . I
1.0 03-]:

Figure 16. PMR spectrum of 5,5'—Bis(3—oxo-propenyl)—2,2'—
dipyrrylmethane (ZR).

 

 

71

 

V—r v l v ' V V T V V V r r V V f V
v v ' ~v ~r—v f -
.

 

gun

§s~
1

c—J—

 

 

 

 

 

 

 

 

1 L 1 A A l A; 1 A
J T 4 I A A A I A l L4 A I +4 4 I J L J A A l4 A A
.o )0 ) so m ' on 10 in 0

Figure 17. PMR spectrum of 2,5-Bis(triphenylmethyl)pYrrole
(1%).
'b

 

fiw—v—r—v fivvlvv v1:‘v~’v.v [vv v vlfvvvlrv V—VIY VV‘VI‘V' ITV—I V
r v v v v v

 

   

 

 

 

 

 

:L I ' : I ' f V I fi— ' I ' ' ' iv
,ocn CH 0 o W "W
, TbsCHZNHsC\ 3 3‘:C8NHCH2T08 w
/ CHZCHZCHZ
i
I
I .
I . ! f '
- f . J I,
I ' .-
I p .
l
u‘gi 1*.;.{ * 1* - *i% r-‘ f-. %.i -;f i i.-- 1 -Ii

 

Figure 18. PMR spectrum of d1methy1—N,N'—Bis(tosylmethyl)-
glutarimidate (%QQ)'

72

 

 

 

9 9 no
TOCCHZNHCCHZCHZCHZCNHCHZTOS

* /

. l
I . .l ..I. ..li...11..11 l 1..AIJ_.

8.0 1.8 80 up "u '9' u Jo 20 1.0 o

 

 

 

 

Figure 19. PMR spectrum of N,N'—Bis(tosylmethy1)g1utaramide
(i8).
’\1

 

 

Toscuzunccuzcnzcuzcuncnz'ro.

. i

 

 

 

 

 

 

 

Figure 20. PMR spectrum of N,N'-Bis(tosylmethyl)thiog1utar—
amide (12)

73

 

p4
.—
J.—

'—*1' I 1

I
l I
5m we no

 

° 9
Hozccuznnccuzcuzcuzcnucuzcozu

 

 

 

 

 

 

1‘1 1 .1.I11I.T...I 1.11_L.
5 10

Figure 21. PMR spectrum N,N'—Bis(g1ycy1)glutaramide (g1).

 

 

I

 

 

 

 

 

\4
NH HN
.
,3. JJ—
JAL A
J l l 1 l
l

 

 

Figure 22. PMR spectrum of 3,U,3',H'-Tetramethyldipyrryl-
2,2'—hexacyclotrimethine tetrafluoroborate (gm

 

7H

 

 

NH HN

 

 

 

 

l A IL .‘L A A 1 A A A A L L A A— A l A A A A L A L A A i
A A A I J A__A A 1 A A 4A A l A L A A I l A A A E A A AAglA A 4A 4—1 4 4 J L r A A J A I
80 70 50 w u... ‘ 40 10 20 10 0

 

Figure 23. PMR spectrum of 1,3—Bis(3,u—dimethylpyrrol-Z-yl)—
cyclohexane (§Z)°

 

 

 

 

 

 

 

 

 

 

 

 

' , , 4' a 1

r 1 f T
3m 4w 0H
III.

1

i

o

3

arm '

I

1

Jkr: +* 1

1'

A A
1 Irv-— v .....v

I

I

J ..... 1111 1-._11I.--.1-4..11
PAAI A_LJLAA [A AIIAAAIALAAIILALAIAAAAIAAA+IJ

In re 50 50 mm 40 30 20 10 0

Figure 2A. PMR spectrum of 1,3-Bis(5-formy1-3,U-dimethy1-
pyrrol—Z-yl)cyclohexane (gQ).

75

 

 

 

 

 

 

 

 

 

 

 

I I L I ‘ ‘ I ' ' ' l 'g 1 I ‘ ' ' L l ' I L‘ l
' I F l T I T r
no 480 300 200 Im on:
mo
1 /‘
I
I m
l A A A L A A A A A
A A A A A A A A4L A L A A 14A A A A LA
80 70 50 so In. " co )0 20 10 0

Figure 25. PMR spectrum of 3,U,8,9,17,18,22,23—Octamethy1—30,-
31,33,3u-tetraazaheptacycloE23.3.1.12’5.17’1O.111’15_

 

 

.116’19.121’2“]tetratriaconta—2,4,6,8,10(33),16,18,-
20,22,2U(30)-decaene (gg).
"I'"fI'"'l'f"l"'v'lfrf71""1'*7'lv"'J‘
I ' ' ' i 7 77‘ ' k ' " '7; ' ’ ' L 7 ' 7 ' {4
q \ r
NH HN

 

 

 

 

t L _L L A A L A A l A A
1 A J A A I J A A A_ I A A A A I A A A A

A A A A A A A A
.0 I0 50 so In. "' no 10 )0 10

 

0“—

Figure 26. PMR spectrum of l,3-Bis(3,U-dimethylpyrrol—Z-y1)-
cyclohex—l—ene (30).

 

76

 

 

. f NH rm

 

goo“?

 

 

 

 

Figure 27. PMR spectrum of l,3-Bis(5—formyl-3,u—dimethylpyrrol-
2—yl)cyclohex-1-ene (3%).

 

 

 

 

 

 

fileffi' I' V'I'V'fIf—Tvvljjfi'L""I""Ifi'l"i‘
1 ' f firfi ' V ' Ii F' I'fifi ‘l ' l
3” am no )0“ 'w Om
"I.
\
T NH H”.
BF4°
3
I
i
1
1 . Al LA 1 A l
A TA AIAALA1.A Al AAAIAAAAI.--
no 70 .0 so m‘“ no

 

 

Figure 28. PMR spectrum of 3,5,3',5'-tetramethyl-U,U'-di-

ethyldipyrryl—2,2'-hexacyclotrimethine tetra—
fluoroborate (3g).

 

  
 

3 03057 7989

ll)1|lVfllHIIWWIIIHWIIIHHWW\HIIHHHIH